pr\u00e9sentent<\/p>\n\n\n\n
Autoris\u00e9 \u00e0 pr\u00e9senter ses travaux en vue de l\u2019obtention du Doctorat de l’Institut Polytechnique de Paris, pr\u00e9par\u00e9 \u00e0 T\u00e9l\u00e9com SudParis en :<\/p>\n\n\n\n
le JEUDI 30 MARS 2023 \u00e0 10h00<\/p>\n\n\n\n
Amphi 5
19 place Marguerite Perey, 91120 Palaiseau<\/p>\n\n\n\n
Membres du jury :<\/strong><\/p>\n\n\n\n M. Sylvain\u00a0LE CORFF<\/strong>, Professeur, T\u00e9l\u00e9com SudParis, FRANCE – Directeur de th\u00e8se Dans ce manuscrit, nous d\u00e9veloppons des architectures d’apprentissage profond pour mod\u00e9liser la consommation \u00e9nerg\u00e9tique et la qualit\u00e9 de l’air de b\u00e2timents. Nous pr\u00e9sentons d’abord une m\u00e9thodologie de bout-en-bout permettant d’optimiser la demande \u00e9nerg\u00e9tique tout en am\u00e9liorant le confort, en substituant au traditionnel simulateur physique un mod\u00e8le num’eriquement plus efficace. A partir de donn\u00e9es historiques, nous v\u00e9rifions que les simulations de ce m\u00e9tamod\u00e8le correspondent aux conditions r\u00e9elles du b\u00e2timent. Cependant, les performances des pr\u00e9dictions sont d\u00e9grad\u00e9es dans certaines situations \u00e0 cause de diff\u00e9rents facteurs al\u00e9toires. Nous proposons alors de quantifier l’incertitude des pr\u00e9dictions en combinant des mod\u00e8les \u00e0 espaces d’\u00e9tat \u00e0 des mod\u00e8les d’apprentissage profond pour les s\u00e9ries temporelles. Dans une premi\u00e8re approche, nous montrons comment les poids d’un mod\u00e8le peuvent \u00eatre affin\u00e9s par des m\u00e9thodes de Monte Carlo s\u00e9quentielles, afin de prendre en compte l’incertitude sur la derni\u00e8re couche. Nous proposons un second mod\u00e8le g\u00e9n\u00e9ratif \u00e0 \u00e9tats latents discrets, permettant une proc\u00e9dure d’apprentissage moins co\u00fbteuse par Inf\u00e9rence Variationnelle ayant des performances \u00e9quivalentes sur une t\u00e2che de pr\u00e9vision de l’humidit\u00e9 relative. Enfin, notre derni\u00e8re contribution \u00e9tend l’utilisation de ces mod\u00e8les discrets, en proposant une nouvelle loi a priori bas\u00e9e sur des ponts de diffusion. En apprenant \u00e0 corrompre puis \u00e0 reconstruire des \u00e9chantillons de l’espace latent, notre mod\u00e8le est capable d’apprendre la distribution a priori, quelle que soit la nature des donn\u00e9es.<\/p>\n\n\n\n In this thesis, we develop deep learning architectures for modelling building energy consumption and air quality. We first present an end-to-end methodology for optimizing energy demand while improving indoor comfort, by substituting the traditionally used physical simulators with a much faster surrogate model. Using historic data, we can ensure that simulations from this metamodel match the real conditions of the buildings. Yet some differences remain, due to unavailable and random factors. We propose to quantify this uncertainty by combining state space models with time series deep learning models. In a first approach, we show how the weights of a model can be finetuned through Sequential Monte Carlo methods, in order to take into account uncertainty on the last layer. We propose a second generative model with discrete latent states, allowing for a simpler training procedure through Variational Inference and equivalent performances on a relative humidity forecasting task. Finally, our last work extends on these quantized models, by proposing a new prior based on diffusion bridges. By learning to corrupt and reconstruct samples from the latent space, our model is able to learn the complex prior distribution, regardless of the nature of the data.<\/p>\n","protected":false},"excerpt":{"rendered":" L’Ecole doctorale : Math\u00e9matiques Hadamard et le Laboratoire de recherche SAMOVAR – Services r\u00e9partis, Architectures, Mod\u00e9lisation, Validation, Administration des R\u00e9seaux pr\u00e9sentent l\u2019AVIS DE SOUTENANCE de Monsieur Max COHEN Autoris\u00e9 \u00e0 pr\u00e9senter ses travaux en vue de l\u2019obtention du Doctorat de l’Institut Polytechnique de Paris, pr\u00e9par\u00e9 \u00e0 T\u00e9l\u00e9com SudParis en : \u00ab M\u00e9tamod\u00e8les et approches bay\u00e9siennes […]<\/p>\n","protected":false},"author":4,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"ocean_post_layout":"","ocean_both_sidebars_style":"","ocean_both_sidebars_content_width":0,"ocean_both_sidebars_sidebars_width":0,"ocean_sidebar":"0","ocean_second_sidebar":"0","ocean_disable_margins":"enable","ocean_add_body_class":"","ocean_shortcode_before_top_bar":"","ocean_shortcode_after_top_bar":"","ocean_shortcode_before_header":"","ocean_shortcode_after_header":"","ocean_has_shortcode":"","ocean_shortcode_after_title":"","ocean_shortcode_before_footer_widgets":"","ocean_shortcode_after_footer_widgets":"","ocean_shortcode_before_footer_bottom":"","ocean_shortcode_after_footer_bottom":"","ocean_display_top_bar":"default","ocean_display_header":"default","ocean_header_style":"","ocean_center_header_left_menu":"0","ocean_custom_header_template":"0","ocean_custom_logo":0,"ocean_custom_retina_logo":0,"ocean_custom_logo_max_width":0,"ocean_custom_logo_tablet_max_width":0,"ocean_custom_logo_mobile_max_width":0,"ocean_custom_logo_max_height":0,"ocean_custom_logo_tablet_max_height":0,"ocean_custom_logo_mobile_max_height":0,"ocean_header_custom_menu":"0","ocean_menu_typo_font_family":"0","ocean_menu_typo_font_subset":"","ocean_menu_typo_font_size":0,"ocean_menu_typo_font_size_tablet":0,"ocean_menu_typo_font_size_mobile":0,"ocean_menu_typo_font_size_unit":"px","ocean_menu_typo_font_weight":"","ocean_menu_typo_font_weight_tablet":"","ocean_menu_typo_font_weight_mobile":"","ocean_menu_typo_transform":"","ocean_menu_typo_transform_tablet":"","ocean_menu_typo_transform_mobile":"","ocean_menu_typo_line_height":0,"ocean_menu_typo_line_height_tablet":0,"ocean_menu_typo_line_height_mobile":0,"ocean_menu_typo_line_height_unit":"","ocean_menu_typo_spacing":0,"ocean_menu_typo_spacing_tablet":0,"ocean_menu_typo_spacing_mobile":0,"ocean_menu_typo_spacing_unit":"","ocean_menu_link_color":"","ocean_menu_link_color_hover":"","ocean_menu_link_color_active":"","ocean_menu_link_background":"","ocean_menu_link_hover_background":"","ocean_menu_link_active_background":"","ocean_menu_social_links_bg":"","ocean_menu_social_hover_links_bg":"","ocean_menu_social_links_color":"","ocean_menu_social_hover_links_color":"","ocean_disable_title":"default","ocean_disable_heading":"default","ocean_post_title":"","ocean_post_subheading":"","ocean_post_title_style":"","ocean_post_title_background_color":"","ocean_post_title_background":0,"ocean_post_title_bg_image_position":"","ocean_post_title_bg_image_attachment":"","ocean_post_title_bg_image_repeat":"","ocean_post_title_bg_image_size":"","ocean_post_title_height":0,"ocean_post_title_bg_overlay":0.5,"ocean_post_title_bg_overlay_color":"","ocean_disable_breadcrumbs":"default","ocean_breadcrumbs_color":"","ocean_breadcrumbs_separator_color":"","ocean_breadcrumbs_links_color":"","ocean_breadcrumbs_links_hover_color":"","ocean_display_footer_widgets":"default","ocean_display_footer_bottom":"default","ocean_custom_footer_template":"0","ocean_post_oembed":"","ocean_post_self_hosted_media":"","ocean_post_video_embed":"","ocean_link_format":"","ocean_link_format_target":"self","ocean_quote_format":"","ocean_quote_format_link":"post","ocean_gallery_link_images":"off","ocean_gallery_id":[],"footnotes":""},"categories":[286,543],"tags":[],"class_list":["post-5907","post","type-post","status-publish","format-standard","hentry","category-fractualites-ennews-fr","category-seminaire-istec","entry"],"_links":{"self":[{"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/posts\/5907","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/users\/4"}],"replies":[{"embeddable":true,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/comments?post=5907"}],"version-history":[{"count":1,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/posts\/5907\/revisions"}],"predecessor-version":[{"id":5908,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/posts\/5907\/revisions\/5908"}],"wp:attachment":[{"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/media?parent=5907"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/categories?post=5907"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/tags?post=5907"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
M. St\u00e9phane\u00a0LECOEUCHE<\/strong>, Professeur, IMT Lille-Douai, FRANCE – Rapporteur
M. V\u00edctor\u00a0ELVIRA<\/strong>, Professeur, University of Edinburgh, ROYAUME-UNI – Rapporteur
Mme Marie-Pierre\u00a0ETIENNE<\/strong>, Ma\u00eetresse de conf\u00e9rences, Institut Agro Rennes-Angers, FRANCE – Examinatrice
M. Wojciech\u00a0PIECZYNSKI<\/strong>, Professeur, T\u00e9l\u00e9com SudParis, FRANCE – Examinateur
M. Maurice\u00a0CHARBIT<\/strong>, Professeur \u00e9m\u00e9rite, Accenta, FRANCE – Co-encadrant de th\u00e8se<\/p>\n\n\n\n
R\u00e9sum\u00e9 :<\/strong><\/p>\n\n\n\n
Abstract : \u00ab\u00a0Metamodel and bayesian approaches for dynamic systems\u00a0\u00bb<\/strong><\/p>\n\n\n\n